THE BOOK--Playing The Percentages In Baseball

Fires and Extinguishings?
Fires and Quenches?
Fires and Hoses? (actually, I like this one, as “hosing” someone kind of means delaying them in some slang senses)
Chokes and Smokes?
Meltdowns and Shutdowns?

I like it a lot.

Jonathan Broxton: 38 and 8
Andrew Bailey: 26 and 7
Brian Wilson: 36 and 11
Jonathan Papelbon: 35 and 3

I notice that if I fiddle with the threshold, the “arsons” usually stay the same, while the “douses” vary.  The “arsons” usually have quite a high value, often in the -0.3 or worse range.

Could someone post numbers for non-closers?

Game...Douses…

Love it.

How about and “Smother” and “Torch”

Non closers:

Jeremy Affeldt: 32 and 10 or (0.1 threshold) 18 and 6
Todd Coffey: 25 and 10 or (0.1 threshold) 13 and 8
David Weathers: 25 and 13 or (0.1) 14 and 10

Smothers, Smotherings.
In a sentence: Yes, Jim. Mariano has smothered the opposition in 15 games this season.

Or moving away from fire terms, hindered, halt, prevent.

I like “Torch”.  Torches.
Smothers sounds funny, though.

I like meltdowns and shutdowns.

What’s the overall average for typical save situations?  One of the two happens more frequently than the other, right?

Rethinking it “smothers” makes me think murderer.

i love it.  would there be a pitch/batter/inning limitation to earn a douse/arson?

I like Meltdowns and Shutdowns.  You can use those in a sentence.  I also like Torches.

dutch: I don’t know why you would put in a limitation.

***

More interestingly: how would an ace reliever react if he’s told that if he enters a game with a 3-run lead to start the 9th inning that he has no chance at having a Shutdown?  Would he start to lobby to pitch in the 8th inning?

***

What kind of record do we want to see for say Mariano Rivera, an average reliever, and a crappy reliever?  That will tell us what kind of threshholds to set.

Uhhh… studes… I’m having a flashback… didn’t you do something really cool with relievers about two years ago in one of your batted ball reports, and I told you that you should write it up somewhere?  I just have this vague memory that I liked whatever you did.

tango - apologies for not explaining myself better.  as per my understanding of wpa you can rack it up by either the quality/leverage of the play or by the number of plays.  (ie a key strikeout with RISP in a 1 run game or a 1-2-3 inning with a slightly larger lead)

i am not advocating a limitation, i was just wondering.  as a mets fan, i immediately think of how feliciano, nieve, and k-rod would fare with this new concept.  stereotypically, feliciano is a 1-2 batter lefty specialist, nieve is a longer reliever, and krod is the 9th inning only guy and i wonder how their roles might affect this stat for them.

sorry to ramble on, but it also gets me thinking of a similar threshold for starters (or even long relievers or 8th inning guys) and how that would shake out.

Isn’t the problem with saves that the definition of “high leverage” as being three runs or less is somewhat arbitrary?

Are we more interested in how the reliever pitched, or are we more interested in how much their performance impacted the game?  I think the traditional save statistic was trying to capture the latter.

@ dutchbrowncoat #16

This year:
Feliciano would be at 4 & 0 so far in 14 appearances
Nieve would be at 5 & 1 in 16 appearances
K-Rod would be at 3 & 1 in 11 appearances

If you want this to catch on anywhere in the mainstream, you’ll want the basic criteria not EXPLICITLY based on WPA, but defined in such a way that it almost always yields the right situations that if you used WPA. If the BPro guys have issues getting the idea behind WPA, square those problems to see how ESPN will get it, and square THAT to see how others will get it.

Charlie, for the moment, I would say this is more important:

“1. ignore if it’s sellable “

Get it right, get it the way you want it, then worry about getting the shortcut version.

In the honor of FJM, how about a food metaphor.

Preserve v Spoil

I suggest “ice” because it makes the other team go cold, and it is only three letters.

Tom - First thanks for taking up the cause.  I am not sure on the level of WPA, but seems to be pretty good.

I thought that gasoline and water may work for the name, but shutdown and meltdowns make the most sense so far.

"Uhhh… studes… I’m having a flashback… didn’t you do something really cool with relievers about two years ago in one of your batted ball reports, and I told you that you should write it up somewhere?  I just have this vague memory that I liked whatever you did.”

Yes, I did an analysis of which relievers had “dramatic” saves by examining how LI moved during their appearance.  If LI moved up during their appearance but they still got a save, that was a dramatic save. K-Rod was the leader in dramatic saves.  I thought it was pretty cool, but the audience was limited and I didn’t write it up again.  I should do that.

Looking at Mariano’s WPA from 2009, the .05 WPA threshold seems to exclude saves that came with 3 run leads and includes games where the score is tied and no runs are allowed.

So if someone wants to come up with a better definition of a save, it doesn’t look like it would be very difficult.

I purposefully chose the .05 to exclude the 3-run saves.

And yeah, keeping the score tied is valuable.

Here is the discussion at RoyalsReview when this new format given.  Results seem a little mixed, but I think it is much better than saves.

http://www.royalsreview.com/2010/5/5/1443967/shutdowns-and-meltdowns#comments

I like this idea a lot.

But the biggest issue here is this:

“And Win Probability doesn’t solve the sticky issue of splitting credit between pitching and fielding”

A big error from a fielder would effectively give that pitcher a meltdown. And having a good defense could realy boost someone’s WPA.

For a primer on WPA, go here:

http://www.hardballtimes.com/main/article/the-one-about-win-probability/

Based on what I see, if a reliever came in during the 6th, 7th, 8th, or 9th and allowed no runs, they would earn a shutdown. All of those added more than +.05

Like I said, YOU set the threshhold after playing around so that you get the results you want.

As for pitching/defense: we are limited to whatever data is recorded.  If this is not an issue for Saves, then it’s not an issue for meltdown/shutdowns.

Sorry, I was just pointing that out since someone said that keeping the score tied is valuable, and I agree that it is. I was just saying that it would qualify for a shutdown if that happened.

It would be interesting to see how many shutdowns a bullpen could get in a game.

As for pitching/defense, I understand we are limited, I was just pointing out a flaw.  It would just suck for a reliever to come in, have a fielder commit an error, and he gets pulled for someone else and getting a nice meltdown to show for it.  And I dont see how it’s not an issue for saves, as a reliever can easily have a save blown due to a fielders error.

I really like the idea overall.

What I mean is that it’s an existing flaw for any metric, and not limited to the xDown metrics.  So, I don’t know that it needs to be discussed as anything other than a limitation (and certainly not a “flaw").

The only reason I bring it up is an added baserunner can swing the WPA by more than -.05. So if there’s a error allowing a baserunner to get on base, then a new reliever comes in to replace the original reliver, the original reliever would get a meltdown. An added baserunner does nothing for a Save/Blownsave unless they score.

Right, taht is a good point.  For a loogy pitcher, one baserunner can turn the situation into a meltdown.

At the same time, a guy hits a ball over the fence that Gutierrez snags for an out.  It’s an out for that hitter, even if he was not really “responsible” for making an out. 

That’s why, we have to get away from worry about proper credit, and simply think in terms of assignments and tracking.

In any case, it’s not like the pitcher was 100% faultless… he did throw that pitch.  And, he has no expectation of 100% clean play from his fielders.  Part of the cost of pitching to contact is that a fielder will muff a play.  That’s part of the deal, and the pitcher who pitches to contact is going to be partly responsible for that.

meltdowns: UBU (unintentional blow up)

If you want this to catch on anywhere in the mainstream, you’ll want the basic criteria not EXPLICITLY based on WPA, but defined in such a way that it almost always yields the right situations that if you used WPA. If the BPro guys have issues getting the idea behind WPA, square those problems to see how ESPN will get it, and square THAT to see how others will get it.

Well, it’s not like I don’t get the idea behind WPA, I just disagree with it being used to answer certain questions. It’s not that I can’t follow the math - I would probably be against any “lite” version of Shutdowns/Meltdowns that avoided using WPA in favor of some more simplistic critera on the same grounds, assuming it was being used to come to much the same result.

Which questions do you like WPA to answer?